During one of the training sessions I was delivering for a group of IT graduates, our conversation moved toward the skills and career paths that are becoming more relevant as artificial intelligence changes the workplace. One of the participants mentioned a field that has become increasingly visible in recent years: prompt engineering.

The term stayed with me, because its emergence says something important about the way technology reshapes work. Not long ago, prompt engineering was hardly part of mainstream conversations about careers. Then generative AI tools became widely accessible, organizations began experimenting with them, new challenges emerged, and a new set of skills started gaining attention.

At almost the same time, we were hearing a very different message: AI is going to take our jobs. I found the contrast interesting. The same technology that creates anxiety about job displacement is also creating new skills, new roles, and new opportunities. We have seen versions of this throughout the history of technology. Computers changed how work was done. The internet created entirely new industries. Smartphones changed consumer behavior and created new business models. AI is accelerating that cycle.

So when we ask, "Will AI take our jobs?", I think we are asking a question that is too narrow. The more useful question is: how will AI change the work we do today?

A Job Can Change Without Disappearing

One of the easiest mistakes to make when discussing AI and employment is to treat a job as one indivisible activity. Most jobs are collections of tasks. Some are repetitive. Some require analysis. Some rely on experience. Others require communication, relationships, creativity, judgment, or accountability.

When AI enters the picture, it can automate some of these tasks without eliminating the entire role. The job remains, but its composition changes. Some activities become faster. Others become less valuable. New responsibilities emerge. The skills required to perform the same role effectively can change significantly.

That leads to a more useful question for professionals: if AI can do a significant part of my job, where does my value come from? That question forces us to think beyond the tasks we currently perform and toward the capabilities we bring to those tasks.

The Prompt Engineering Example Tells Us Something Bigger

When one of the graduates mentioned prompt engineering, I did not focus on whether prompt engineering itself would remain a distinct profession. I became more interested in what its emergence tells us about the labor market.

A new technology creates new interactions. Those interactions create new problems. Those problems create demand for new capabilities. Generative AI created a need for people who understand how to communicate effectively with AI systems, structure instructions, evaluate outputs, and integrate these tools into real workflows.

Some of these skills may eventually become standard parts of other professions. The job title may evolve. The skill itself may become expected rather than specialized. That is perfectly normal.

The bigger lesson is that technology removes some needs while creating others. We tend to focus heavily on the first part. We should pay much more attention to the second.

The Real Change May Happen Inside the Job

A professional may keep the same job title while doing a very different job five years from now. Consider someone who spends hours researching information, organizing it, summarizing it, and preparing an initial report. AI can now reduce the time required for many of these activities dramatically.

That does not automatically make the professional irrelevant. It changes where their time can go. They can spend more time interpreting the information, testing assumptions, speaking with stakeholders, exploring alternatives, identifying risks, or thinking about the actual problem the report is supposed to solve.

The value shifts from producing information to understanding it. It shifts from completing tasks to exercising judgment. It shifts from creating outputs to knowing what those outputs should be used for. That is a much more interesting way to think about the future of work.

Experience Still Matters, But Its Value Is Changing

There is a tendency to assume that AI will make human experience less valuable. I see something different happening. Experience may become more valuable precisely because information is becoming easier to access.

For a long time, an experienced professional had an advantage because they had accumulated knowledge and could reach an answer faster than someone with less experience. AI changes part of that equation. A machine can now retrieve and organize enormous amounts of information in seconds. Knowing the information is therefore only one part of the value.

The more valuable capability becomes understanding context, asking the right question, evaluating the answer, spotting what is missing, and deciding what to do next. AI can give you ten possible answers. Experience helps you understand which one deserves attention. That makes professional judgment increasingly important.

AI Should Make Us Think Better, Not Think Less

One of my concerns about AI adoption is the possibility that people will use it to avoid thinking rather than to improve their thinking. We can now ask AI to create a strategy, prepare a presentation, analyze data, write a report, or draft a proposal in minutes. But producing an answer quickly does not mean we have solved the right problem.

A strategy can be polished and still address the wrong issue. A report can be well structured and still contain weak assumptions. A presentation can look convincing while missing what the audience actually needs. AI can make weak thinking look professional. That makes human judgment more important, not less.

The people who create the most value will be those who can question the output, challenge assumptions, verify information, and connect the answer to the real context in which a decision needs to be made.

A Good Prompt Still Needs Good Thinking

Prompt engineering is an interesting example because it can easily be misunderstood as a technical shortcut. A well-designed prompt can certainly improve the quality of an AI output. But the quality of the prompt is closely connected to the quality of the thinking behind it.

If you understand the problem, you know what to ask. If you understand the field, you know what to look for. If you have experience, you know what deserves skepticism. If you can evaluate results, you know what can be used and what needs further validation.

That is why I believe the future will favor professionals who combine genuine domain expertise with the ability to work effectively with AI. A developer who knows how to use AI in software development will work differently from one who ignores it. A designer who integrates AI into experimentation and ideation will approach the creative process differently. A manager who understands how AI can support decision-making will lead differently from one who sees it only as a productivity tool.

The goal is not for everyone to become an AI specialist. The goal is for people to understand how AI can expand their own professional capabilities.

Managers Will Have to Rethink Their Role

The change will reach management as well. Managers who spend significant time assigning tasks, monitoring details, reviewing routine work, and answering repetitive questions may find that some of these activities become easier with AI. That creates an opportunity.

Managers can spend more time developing people, creating clarity, improving decisions, setting priorities, and helping teams think more effectively. We may gradually move from management that focuses heavily on monitoring execution toward management that focuses more on the quality of thinking and decision-making.

That requires a different kind of leadership. Managers will need enough technological understanding to work effectively with AI, but they will also need a deep understanding of people, context, culture, and capability building.

Organizations Need Capability Building, Not Just AI Tools

It is easy for an organization to buy AI tools, run workshops, and announce an AI transformation initiative. Real transformation requires a deeper question. What do we want our people to become capable of doing better because of AI?

That question changes the conversation. Instead of starting with technology, we start with value. Which tasks should be automated? Which processes should be redesigned? Which capabilities need to be developed? Which roles will change? Where will human judgment remain critical?

These questions can also reveal something uncomfortable. Some processes we have spent years trying to optimize may need to be questioned rather than optimized. Sometimes the right answer is not to add AI to the process. It is to ask why the process exists in the first place.

Technology Redistributes Value

Every major technological shift changes where value sits in the labor market. Some skills become easier to access. Some tasks become faster and cheaper. Other capabilities become more valuable. AI is doing this at a remarkable speed.

When producing a first draft becomes easier, knowing what the message should be becomes more valuable. When basic data analysis becomes faster, knowing which questions to ask becomes more important. When information becomes easier to access, knowing how to evaluate and apply it becomes more valuable.

Technology does not simply remove value. It changes where value is created. That may be one of the most important ideas for professionals thinking about their future.

What Should Graduates Learn Today?

I keep coming back to that conversation with the IT graduates because it captures a challenge that many young professionals are facing. A graduate entering the workforce today is entering a very different environment from someone who started their career a decade ago. Knowledge is more accessible. Tools are more powerful. Expectations are changing faster.

A university degree remains important, but it cannot be the end of the learning journey. Graduates need to build capabilities that allow them to adapt. Understanding technology matters. Understanding problems matters more. Learning tools matters. Knowing how to keep learning may matter even more.

Instead of asking only, "What job is in demand today?", I would encourage graduates to ask another question: "What capabilities will continue to create value even when the jobs themselves change?" That is a much stronger foundation for a career.

Perhaps We Need a Different Question

After that conversation with the graduates, I started thinking differently about the way we discuss AI and employment. We keep asking which jobs AI will take. We should also ask which skills, roles, and opportunities AI will create.

AI will automate some tasks. It will change some jobs. It will create new ones. That cycle will continue as the technology evolves. The bigger opportunity may exist in the space between human capability and technological capability.

There are things AI can do faster and at greater scale. There are things that require human judgment, creativity, relationships, context, accountability, and responsibility. The opportunity lies in combining the two.

So the question I would rather ask employees, graduates, managers, and organizations is this: if AI can do a significant part of what I do today, what value will I be able to create tomorrow? That question is more uncomfortable than asking whether AI will take our jobs. It is also more useful.

The labor market will continue moving whether we are ready or not. Skills will change. Roles will change. New opportunities will emerge that we cannot clearly define today. We do not need to predict every future job. We need to build the ability to learn, adapt, and rebuild our capabilities as the nature of work changes.

Perhaps that is the most important lesson I took away from that conversation with the graduates: AI may take some tasks away from people, but it can also create the space for people to do work that is more valuable. The real question is what we choose to do with that space.